Atrial fibrillation (AF) is the most common sustained arrhythmia worldwide and it is a known risk factor for stroke. Despite a much higher prevalence of stroke in African-Americans compared to Whites, African Americans have a lower prevalence of AF. African-Americans also experience high rates of risk factors associated with AF including hypertension, obesity and diabetes. It has been theorized that this paradox of unexpected low rates of AF in African Americans may be due to differential detection of AF by race (ascertainment bias). Most prior studies have defined AF by self-report, hospitalizations, claims data, and/or 12-lead electrocardiography, all of which may miss subclinical and intermittent AF. The detection of AF can be challenging as it can be paroxysmal and asymptomatic requiring ambulatory electrocardiography (Holter) for ascertainment. The National Heart Lung and Blood Institute (NHLBI) has called for advancements in AF epidemiology through increased surveillance of AF in longitudinal studies, especially among non-White ethnic groups. We propose to measure the race-specific prevalence of AF, in an informative study nested in the ongoing, bi-ethnic Atherosclerosis Risk in Communities (ARIC) Study. With the goal of estimating the prevalence and characteristics of AF by race, we plan to take advantage of the ongoing 5th ARIC cohort examination (June 2011-May 2013) inclusive of state-of-the-art echocardiograms (ECHOs) as a way to efficiently identify an informative subgroup for Holter monitoring. This study has four main aims. Aim 1: Estimate the race-specific prevalence of AF in African-Americans and Whites aged 70 years and older. Aim 2: Calibrate the prevalence of AF by quantifying under-ascertainment in the detection of AF according to length of Holter monitoring. Aim 3: Evaluate the proximal risk factors and triggering factors of AF in African Americans from the electrocardiographic, echocardiographic, spirometry, and biomarker measures performed proximal to Holter monitoring. Aim 4: Identify the metabolic, behavioral and comorbid factors in middle age, and their change over the course of the subsequent four cohort examinations, associated with the incidence of AF in African Americans. We will perform a staged validation of indicators of AF, and 48-hour continuous Holter monitoring on participants without substantiated AF, for 825 African American and 400 White members of the ARIC cohort from 2 study sites. Stratified random sampling will be based on race, center, gender, and characteristics that increase yield of AF(enlarged left atrial size or low ejection fraction on echocardiogram, or clinical heart failure). From the prior characterization of the ARIC cohort's 5 examinations and over 25 years of comprehensive follow-up for events, we will be able to efficiently examine the race-specific prevalence and characteristics of AF, with safeguards against under-ascertainment not available in published reports. This study will investigate the racial paradox of low atrial fibrillation in African Americans which has implications for the prevention and clinical management of AF in African Americans.